p0.95 & p0.96 find no PGP-encrypted content

p0.95 & p0.96 find no PGP-encrypted content

[Karl Kleinpaste]

I've just discovered that 0.95 and 0.96 are unable to decrypt any of
my mail containing PGP-encrypted messages.

I have mailcrypt 3.5.4 and PGP 5.0i (from replay.com's
pgp-5.0i-7.i386.rpm).  However, considering that there is no
*MailCrypt* buffer ever created, and that the minibuf complaint is
"Found no encrypted message in this buffer," I tend to think that 0.9x
has done Something Bad.  Anyone see similar misbehavior, or have
another explanation for what I see?

Well, hang on a minute...

I just tried edebug-defun through mc-decrypt-message, which in turn
led me to edebug-defun on mc-message-delimiter-positions.  This is
most odd: The failure is in not finding the PGP bound delimiter
strings via re-search-forward.  When I execute re-search-forward on my 
own with the exact same arguments, it comes up with the correct buffer 
position, but when run as part of the mailcrypt functions, it returns
nil.

I don't get it.  Is this an XEmacs 21.2 beta bug of some weird sort?

--karl

Re: p0.95 & p0.96 find no PGP-encrypted content

[Lars Magne Ingebrigtsen]

Karl Kleinpaste <karl@justresearch.com> writes:

> I just tried edebug-defun through mc-decrypt-message, which in turn
> led me to edebug-defun on mc-message-delimiter-positions.  This is
> most odd: The failure is in not finding the PGP bound delimiter
> strings via re-search-forward.  When I execute re-search-forward on my 
> own with the exact same arguments, it comes up with the correct buffer 
> position, but when run as part of the mailcrypt functions, it returns
> nil.

Could it be some `case-fold-search' buggage?  I'm not closely familiar 
with Mailcrypt, so I'm just guessing?

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen

Re: p0.95 & p0.96 find no PGP-encrypted content

[Karl Kleinpaste]

Karl Kleinpaste <karl@justresearch.com> writes:
>> I just tried edebug-defun through mc-decrypt-message, which in turn
>> led me to edebug-defun on mc-message-delimiter-positions.  This is
>> most odd: The failure is in not finding the PGP bound delimiter
>> strings via re-search-forward.  When I execute re-search-forward on my 
>> own with the exact same arguments, it comes up with the correct buffer 

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:
> Could it be some `case-fold-search' buggage?  I'm not closely familiar 
> with Mailcrypt, so I'm just guessing?

This also came up on xemacs-beta.  It's a weird one, I've found: I can
stop the bug merely by manually defun'ing mc-gnus-decrypt-message
(from mc-toplevel.el); after doing so, the error stops occurring.  I
have wondered if there is a byte-compiler error here, but no interest
in this question was found when I raised it there.

Re: p0.95 & p0.96 find no PGP-encrypted content

[Lars Magne Ingebrigtsen]

Karl Kleinpaste <karl@justresearch.com> writes:

> This also came up on xemacs-beta.  It's a weird one, I've found: I can
> stop the bug merely by manually defun'ing mc-gnus-decrypt-message
> (from mc-toplevel.el); after doing so, the error stops occurring.  I
> have wondered if there is a byte-compiler error here, but no interest
> in this question was found when I raised it there.

I've seen stuff like that, but upon inspection, I've always found that 
I had another definition of the function in question squirreled away
in some obscure file or other.  :-)

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen